On 11/2/2016 9:22 AM, Blower, Melanie wrote:
Yes that's a good idea. I was trying to get boost Windows regression testing going awhile ago using Cygwin but I ran into some obstacles, then ran out of time. I hope to get back to that task at some point. Meanwhile would anyone be available to do the testing, Intel could supply a licensed compiler.
Needless to say if Intel supplied a licensed compiler on Windows for developers who would be willing to test it with their Boost library(s) that would help solve problems which you are reporting. Whereas reporting problems with Intel C++ 17.0 on Windows, when it would ne too costly for Boost library developers to buy a license just to test the compiler, is not likely to lead to a solution without the developer being able to use the actual product.
I haven't yet found a way to supply compute resources to go with the license.
Best regards, Melanie Blower
-----Original Message----- From: Boost [mailto:boost-bounces@lists.boost.org] On Behalf Of Edward Diener Sent: Tuesday, November 1, 2016 10:00 PM To: boost@lists.boost.org Subject: Re: [boost] [dll] trouble building boost dll with Intel compiler on Windows
On 11/1/2016 6:55 PM, Gavin Lambert wrote:
On 2/11/2016 11:50, Edward Diener wrote:
If Boost developers could get a free version of Intel C++ on Windows, purely for the non-commercial task of testing Boost with it, it would be easier for them to fix things in libraries in order to get it to work. Although Intel C++ on Linux graciously offers a free version for Boost developers, none is offered for Intel C++ on Windows.
Perhaps Intel could host a Boost test runner for it? While that'd have a longer cycle time for fixes, it should alleviate any licensing concerns they might have.
That would not help much if a developer needs to find out why something is not working when compiled with Intel C++ on Windows. Of course it helps for testing.